Is it better than mp3 in terms of quality, especially in comparison to the --alt-presets?
I assume that the GT3b2 version of Vorbis produces similar filesizes in q5 or q6 (compared to lame 3.90.3 APS). At this level, which encoder does a better job (in quality / bit terms )

You use Ogg Vorbis. What q-setting do you prefer?
I don't want no final wisdom on this, but your honest personal opinion on what setting seems reasonable.

Background:
I have switched from APE to FLAC for backing up my audio-cds (--> DVD-R). Reasons are a better portability and more robust filestructure as far as I can judge. I don't have infinite amounts of time on my hands, so I have to rely on public testing and the author's documents. FLAC is well-documented and true open-source; I like that. I will miss those APLs, but you can't have anything in life...

Anyway, I will keep lossy copies of the individual tracks on my HDD and decided for Ogg Vorbis (reasons same as above). I used Lame APS for a long time and now I ask myself: Is Ogg better? The whole free-for-everyone-aspect appealed to me but now I wonder what compiler, quality-level etc. to use.

Can you help shed some light on this? Thanks in advance for your replies!

I'm not too familiar with lame APS since I haven't used it before so I won't comment on it and how it compares. But I think GT3b2 at q 6 is pretty good. Lossless stereo coupling and less of the HF boost problem. At these sorts of bitrates, it will take a bit of effort to notice any quality flaws.



Personally, I've ripped my CDs using GT3b2 at q 5. However, I'm considering switching to aoTuV at q 5 soon. GT3b2 does have the HF boost problem so aoTuV is expected to be better in this regard. I'm not too sensitive to pre-echo at these high bitrates so it's good for me.



The only thing you lose with switching from mp3 to Vorbis is hardware compatibility. There are only a few hardware players which support Vorbis while nearly all of them support mp3.

No-one has reported any quality problems from binaries compiled with other compilers like mingw32 (gcc), Intel C++ compiler, etc. While bit-wise comparisons reveal differences due to different optimisation techniques, I don't think they are audible differences. So it should be safe to use any one of them. If in doubt, go for an MSVC compile.

As for quality levels, it depends on how good your ears are at noticing artifacts, what level of quality you expect (transparent or just good enough?), and what sort of artifacts you can tolerate and which you consider annoying, etc. Vorbis is a lossy coder so artifacts always exist. It is whether they are audible, to what level of audibility, how irritating they are, etc. which determine the success of it.

I personally recommend GT3b2 or aoTuV at q 5. But the best way is always to listen to the files yourself and make a decision.

Roberto's last multiformat public listening test showed Vorbis to be better than Lame at 128k bits. Unfortuntately, there is no theorem that says the codec that sounds better at 128k bits will sound better at higher bit rates. It is also very difficult to get meaningful tes results as the bit rates go up. Only a few people can hear the difference in abx testing.

The best thing is to listen to both and pick the one you want.

However, if compatibility is important, mp3 wil be the winner. Nearly every portable supports MP3 from $50 cd/mp3 players on up. Vorbis is only on a few premium priced devices and has noticably shorter battery life. Windows XP displays ID3 tags in mP3 files natively. Several CD burning programs (including Nero with a free plugin) support Vorbis, but it does not enjoy the near universal support that MP3 has in all sorts of audio software.

My point of view is that achieving the same degree of transparency at a bit rate that is only 10 to 20 percent less does not make that much difference when portables are available with large hard drives.

http://wiki.xiph.org/VorbisHardware

Quite a few, don't you think?

Not really. Saisonic sells a CD based one in Korea for around 130 dollars.

IMO $130 is premium priced when closeout (Panansonic at Fry's) cd/mp3 players can be had for $30.

As to what "only a few" means as compared to "quite a few" it is in the eye of the beholder.

Look, I don't have anythg against Ogg Vorbis. To that I would add that I appreciate what QuantumKnot and the others are doing with their efforts to tune the best possible 128k bit performance out of Ogg Vorbis. It raises the bar for everyone. That is beneficial no matter what codec you use. If Ogg Vorbis was to place very strongly in the next public listenig test, that would shake a few trees.

Well...let me say that I was a mp3 user and now I use Vorbis for all my portable stuff. I recently re-ripped all my CD's to OGG deleted all my mp3 files.


I use q 7 for all of mine. I have listened to them on my computer speakers, with headphones, on my home stereo, and in my car and they sound great to me (no flaws jump out.) From what others have said I probably could have gone as low as q 5 without issues, but since some of my Vorbis files are my only backup copy, q 7 was the best combination (for me) of file size and sound preference.

I still keep my "important" files in FLAC format and use Vorbis for portability and for not-so-important files. In my humble opinion, Vorbis and FLAC are an awesome combination. We just need more hardware player support for them.

BTW, all of my files are now encoded with GT3b2.